INDEX

Explanations

July 02, date and Brighton

The neuron activates on words that begin with a capital letter (e.g. proper nouns or section headings).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 that

-1.50

 italianos

-1.49

 ardından

-1.47

 kanyang

-1.42

 then

-1.38

 complicado

-1.35

 creativo

-1.31

鹩

-1.31

 дополнитель

-1.27

 decorativas

-1.27

POSITIVE LOGITS

in

1.74

1.43

 have

1.32

 when

1.32

 There

1.31

ellingham

1.31

two

1.28

Only

1.27

時的

1.26

During

1.24

Activations Density 0.003%